In a method for the management of tasks in a decentralized data network with a plurality of nodes for carrying out the tasks, resources are distributed based on a mapping rule, in particular a hash function. A task that is to be suspended is distributed by dividing the process image of the task into segments and by distributing the segments over the nodes using the mapping rule. Thus, a distributed swap space is created so that tasks can also be carried out on nodes whose swap space is not sufficient on its own. The method can be used in embedded systems with a limited storage capacity and/or in a voltage distribution system, wherein the nodes are, for example, switching units in the voltage distribution system. The method can also be used in any other technical systems such as, for example, a power generation system, an automation system and the like.
Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A method for managing computing processes in a decentralized data network comprising a plurality of network nodes for executing the processes, with resources being distributed in the data network based on a mapping rule, wherein the mapping rule may include a hash function, the method comprising: a) stopping one or more computing processes executed on a network node and to be suspended and creating a process image for each stopped computing process; b) breaking down the process image of a respective computing process into slices; and c) distributing the slices of the process image of the respective computing process with the aid of the mapping rule to the network nodes, which creates a distributed process image.
In a decentralized network of computers, this method manages tasks by distributing them across multiple machines. When a running task needs to be paused, the system creates a snapshot of its current state (a "process image"). This snapshot is then divided into smaller pieces ("slices"). These slices are distributed across the network's computers using a mapping rule, which can be a hash function. This creates a distributed version of the task's state, allowing it to be resumed later, potentially on a different machine.
2. The method according to claim 1 , wherein the process image in step b) is split up into slices of the same size.
The method for managing computing processes in a decentralized data network as described, involves breaking down the process image into slices of equal size. This ensures uniform distribution and simplifies reconstruction when the task needs to be resumed.
3. The method according to claim 1 , wherein the slices distributed in step c) are stored in respective main memories in the network nodes.
In the method for managing computing processes in a decentralized data network as described, the slices of the process image, once distributed, are stored in the main memory (RAM) of the network nodes. This allows for faster access and quicker task resumption compared to storing them on disk.
4. The method according to claim 1 , wherein the method is used in a data network in which each network node is responsible for a pre-specified quantity of hash values able to be generated by the hash function.
The method for managing computing processes in a decentralized data network as described, works within a network where each computer is responsible for a specific range of hash values. This means the hash function used to distribute task slices determines which computer stores which slice based on the slice's hash value. This ensures consistent and predictable data distribution.
5. The method according to claim 4 , wherein the data network is a peer-to-peer network based on a distributed hash table, in which a range of hash values is divided up into hash value intervals and each network is responsible for a hash value interval.
This method uses a peer-to-peer network where data is organized using a distributed hash table, meaning the network divides up data responsibilities based on specific ranges of data identifiers.
6. The method according to claim 4 , wherein in step c) a keyword unique in the decentralized data network is generated for each slice of the process image of a respective computing process, with the keyword being mapped with the hash function to a hash value and the slice, for which the keyword was generated being stored in the network node which is responsible for the hash value to which the keyword was mapped.
In the method for managing computing processes in a decentralized data network where each network node is responsible for a pre-specified quantity of hash values able to be generated by the hash function, each slice of a task's snapshot gets a unique identifier (keyword). This keyword is then hashed, and the resulting hash value determines which computer in the network will store that particular slice, ensuring that slices are distributed consistently according to the hash function.
7. The method according to claim 6 , wherein a keyword for a slice of a process image of the respective computing process is created from information about the respective computing process and an identification for the slice, wherein the identification can be a slice number.
In the method for managing computing processes in a decentralized data network as described where each slice of a task's snapshot gets a unique identifier (keyword), the keyword for each slice is generated using information about the task itself (e.g., process ID) combined with a slice identifier (e.g., slice number). This ensures uniqueness across the network and allows for easy reassembly of the task's snapshot.
8. The method according to claim 7 , wherein at least one of the information about the respective computing process and the keyword is stored in the network node which has executed the relevant computing process before it was stopped, and/or in which at least one of the information about the relevant computing process and the keyword is managed by a process management method.
In the method for managing computing processes in a decentralized data network as described, where the keyword for each slice is generated using information about the task itself combined with a slice identifier, either the task information or the generated keyword (or both) is stored on the original computer that was running the task before it was paused. This information can also be managed by a separate process management system. This helps with locating the slices when the task needs to be resumed.
9. The method according to claim 7 , wherein the information about the respective computing process comprises at least one of a process identification of the relevant computing process and an identification of the process image of the respective computing process.
In the method for managing computing processes in a decentralized data network as described, where the keyword for each slice is generated using information about the task itself combined with a slice identifier, the task information includes the task's process ID or an identifier for the entire snapshot of the task. This information helps identify and locate the correct slices when resuming the task.
10. The method according to claim 4 , wherein in step c) a keyword unique in the decentralized data network is generated for each slice of the process image of a respective computing process, with the keyword being mapped with the hash function to a hash value and the slice, for which the keyword was generated being stored in the network node which is responsible for the hash value to which the keyword was mapped, wherein the process based on a distributed process image is resumed by a network node intended for the resumption of the process by the following steps: i) finding and storing the slices of the distributed process image distributed to the network nodes with the aid of the mapping rule, wherein the mapping rule may include the hash function, in the network node intended for resumption of the process; and ii) combining the slices into the process image and starting the process based on the process image in the network node intended for resumption, and wherein a respective slice of the distributed process image is found in step i) by the keyword of the respective slice being mapped with the hash function to a hash value and, based on the hash value, the network node being found on which the respective slice is stored.
In the method for managing computing processes in a decentralized data network where each network node is responsible for a pre-specified quantity of hash values able to be generated by the hash function, a process is paused and its snapshot is divided into slices, each assigned a unique keyword. These slices are distributed across the network based on the hash of their keywords. To resume the process, the target computer finds the slices by rehashing their keywords to locate the nodes where the slices are stored. It retrieves and combines these slices to rebuild the original snapshot, and then restarts the process from that restored state.
11. The method according to claim 1 , wherein the process based on a distributed process image is resumed by a network node intended for the resumption of the process by the following steps: i) finding and storing the slices of the distributed process image distributed to the network nodes with the aid of the mapping rule, wherein the mapping rule may include the hash function, in the network node intended for resumption of the process; and ii) combining the slices into the process image and starting the process based on the process image in the network node intended for resumption.
In the method for managing computing processes in a decentralized data network as described, a running task can be resumed on any node. The node locates all slices of the task's distributed state (process image) across the network using a mapping rule, potentially a hash function. The node then retrieves these slices and reconstructs the original task state. Finally, the node restarts the task from the reconstructed state.
12. The method according to claim 11 , wherein the network node intended for resumption of the process is the same network node that has executed the process before it was stopped.
In the method for managing computing processes in a decentralized data network where a running task can be resumed, the task is resumed on the same computer that was originally running it before it was paused. This ensures that the task resumes in a familiar environment.
13. The method according to claim 11 , wherein the network node intended for resumption of the process is a different network node from the one that has executed the process before it was stopped.
In the method for managing computing processes in a decentralized data network where a running task can be resumed, the task is resumed on a *different* computer than the one that was originally running it before it was paused. This enables load balancing and fault tolerance, allowing tasks to be moved to available resources.
14. The method according to claim 11 , wherein the process based on the process image is started in step ii) such that the assembled process image is loaded into a memory of the network node intended for resumption of the process and is subsequently executed in the main memory.
In the method for managing computing processes in a decentralized data network as described, when a task is resumed, the assembled snapshot of the task is loaded into the target computer's memory. The task then restarts execution from this memory location, effectively restoring it to its previous state.
15. The method according to claim 1 , wherein the method is used in the data network of a technical system with a plurality of technical components, with at least a part of the technical components representing a network node of the data network, respectively.
The method for managing computing processes in a decentralized data network is used within a technical system containing multiple interconnected components. Some of these components act as individual computers (nodes) within the decentralized data network.
16. The method according to claim 15 , wherein the technical system comprises an energy distribution network, wherein the energy distribution network may be an energy distribution substation, with the technical components comprising switching units in the energy distribution network.
The method for managing computing processes in a decentralized data network, implemented within a technical system, is applied to an energy distribution network, potentially an energy distribution substation. In this context, the individual components acting as nodes are the switching units within the power grid.
17. The method according to claim 15 , wherein the technical system comprises an energy generating system, wherein the energy generating system may be an energy generating system based on turbines.
The method for managing computing processes in a decentralized data network, implemented within a technical system, is applied to an energy generating system, potentially a turbine-based power plant. This suggests the individual components control or monitor aspects of energy generation.
18. The method according to claim 15 , wherein the technical system comprises an automation system, wherein the automation system may be a production line.
The method for managing computing processes in a decentralized data network, implemented within a technical system, is used in an automation system, specifically a production line. This suggests the individual components manage or control different stages of the manufacturing process.
19. A decentralized data network embodied in a combination of hardware and software, and comprising a plurality of network nodes including processors for executing computing processes, with resources in the data network being distributed in the network on the basis of a mapping rule, and with the data network being configured such that the computing processes causes: a) one or more computing process to be executed by a processor on a network node and to be suspended in each case is or are stopped and a process image is created for each stopped computing process; b) the process image of a respective computing process is broken down into slices; c) the slices of the process image of the respective computing process are distributed with the aid of the mapping rule to the network nodes, which creates a distributed process image.
A decentralized network of computers, built using both hardware and software, is designed to manage tasks across multiple machines. Each computer runs its own processes. If a task is stopped, the system creates a snapshot of its state (process image), divides it into smaller parts (slices), and distributes the slices to different machines across the network using a defined mapping rule. This results in the task's state being spread across the network.
20. A decentralized data network according to claim 19 , wherein the process image is split up into slices of the same size.
The decentralized data network for managing computing processes as described, splits up the process image into slices that are all the same size. This simplifies management and distribution of the task data across the network.
21. A decentralized data network according to claim 19 , wherein the mapping rule comprises a hash function.
The decentralized data network for managing computing processes as described, uses a hash function as the mapping rule to determine where the slices of a paused task are stored. This ensures consistent and predictable data distribution across the network.
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
December 22, 2008
July 23, 2013
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.